Application of an improved LWR method to real-time aircraft parameter identification problems

In this paper a recently introduced time domain-based parameter identification (PID) method is considered for the estimation of the aerodynamic coefficients of a specific aircraft. The basic method is the local weighted regression (LWR) technique and it is improved through the addition of a data retention-and-deletion strategy. The resulting PID algorithm was tested using NASA F/A-18 HARV flight data for evaluating its capabilities for online real time PID purposes. The results were then compared with theoretical estimates from the wind tunnel analysis and with the PID results from the maximum likelihood method, a well known PID method for off-line applications.